A Machine Learning Approach in Autism Spectrum Disorders: From Sensory Processing to Behavior Problems

Heba Alateyat; Sara Cruz; Eva Cernadas; María Tubío-Fungueiriño; Adriana Sampaio; Alberto González-Villar; Angel Carracedo; Manuel Fernández-Delgado; Montse Fernández-Prieto

doi:10.3389/fnmol.2022.889641

A Machine Learning Approach in Autism Spectrum Disorders: From Sensory Processing to Behavior Problems

Front Mol Neurosci. 2022 May 9:15:889641. doi: 10.3389/fnmol.2022.889641. eCollection 2022.

Authors

Heba Alateyat¹, Sara Cruz², Eva Cernadas¹, María Tubío-Fungueiriño^{3

4

5}, Adriana Sampaio⁶, Alberto González-Villar⁶, Angel Carracedo^{3

7

8

9}, Manuel Fernández-Delgado¹, Montse Fernández-Prieto^{3

5

8

9}

Affiliations

¹ Centro Singular de Investigación en Tecnoloxías Intelixentes da USC (CiTIUS), Universidade de Santiago de Compostela (USC), Santiago de Compostela, Spain.
² The Psychology for Positive Development Research Center, Lusíada University-North, Porto, Portugal.
³ Genomics and Bioinformatics Group, Centre for Research in Molecular Medicine and Chronic Diseases (CiMUS), Universidade de Santiago de Compostela (USC), Santiago de Compostela, Spain.
⁴ Grupo de Medicina Xenómica, Universidade de Santiago de Compostela (USC), Santiago de Compostela, Spain.
⁵ Fundación Instituto de Investigación Sanitaria de Santiago de Compostela (FIDIS), Santiago de Compostela, Spain.
⁶ Psychological Neuroscience Lab, Centro de Investigação em Psicologia, School of Psychology, University of Minho, Campus de Gualtar, Braga, Portugal.
⁷ Fundación Pública Galega de Medicina Xenómica, Servicio Galego de Saúde (SERGAS), Santiago de Compostela, Spain.
⁸ Grupo de Medicina Xenómica, U-711, Centro de Investigación en Red de Enfermedades Raras (CIBERER), Universidade de Santiago de Compostela (USC), Santiago de Compostela, Spain.
⁹ Grupo de Genética, Instituto de Investigación Sanitaria de Santiago (IDIS), Santiago de Compostela, Spain.

Abstract

Atypical sensory processing described in autism spectrum disorders (ASDs) frequently cascade into behavioral alterations: isolation, aggression, indifference, anxious/depressed states, or attention problems. Predictive machine learning models might refine the statistical explorations of the associations between them by finding out how these dimensions are related. This study investigates whether behavior problems can be predicted using sensory processing abilities. Participants were 72 children and adolescents (21 females) diagnosed with ASD, aged between 6 and 14 years (M = 7.83 years; SD = 2.80 years). Parents of the participants were invited to answer the Sensory Profile 2 (SP2) and the Child Behavior Checklist (CBCL) questionnaires. A collection of 26 supervised machine learning regression models of different families was developed to predict the CBCL outcomes using the SP2 scores. The most reliable predictions were for the following outcomes: total problems (using the items in the SP2 touch scale as inputs), anxiety/depression (using avoiding quadrant), social problems (registration), and externalizing scales, revealing interesting relations between CBCL outcomes and SP2 scales. The prediction reliability on the remaining outcomes was "moderate to good" except somatic complaints and rule-breaking, where it was "bad to moderate." Linear and ridge regression achieved the best prediction for a single outcome and globally, respectively, and gradient boosting machine achieved the best prediction in three outcomes. Results highlight the utility of several machine learning models in studying the predictive value of sensory processing impairments (with an early onset) on specific behavior alterations, providing evidences of relationship between sensory processing impairments and behavior problems in ASD.

Keywords: autism spectrum disorders; behavior problems; machine learning; regression; sensory processing.